The same volume, fundamentally different economics
Two vertical SaaS companies. Both serve the same general market. Both process approximately $8M monthly in payment volume today. Their annual payment economics differ by $1.1M — not because one has better technology, more customers, or faster growth. Because one made the payments architecture decisions in the right sequence and the other didn't.
This is not a hypothetical. It is a pattern that plays out across vertical SaaS companies at every volume level. The gap is architectural, not operational. And it is recoverable — but recovery costs 12–18 months and $200,000–$400,000 in build investment that could have been avoided.
Company A: the standard path
Company A launched payments on Stripe at $800K monthly processing volume. The decision was straightforward: fast integration, good developer experience, no compliance overhead, live in six weeks. Stripe's 2.9% flat rate was an acceptable cost at the time.
The company grew. At $3M monthly, a board member asked about payments monetization. The team ran the numbers for the first time: at 2.9% Stripe rate versus a theoretical interchange-plus program, the annual gap was approximately $480,000. A legitimate concern but not yet a crisis.
At $5M monthly the same calculation showed $780,000 annual gap. Migration conversations started. The team discovered the contract had a 6-month notice provision, the integration was tightly coupled to Stripe's specific API response structures, and the compliance program needed for a direct bank relationship didn't exist. The migration estimate: 14 months, $280,000 in build investment.
At $8M monthly — where Company A is today, 14 months after beginning migration — the direct program is live. The migration cost $280,000 and 14 months of economics gap at an average of $55,000 per month: $770,000 in economics foregone during migration. Total cost of the Stripe decision: approximately $1.05M in investment plus foregone economics, not counting the 14 months of team distraction.
Company B: the architecture-first path
Company B was at $1.2M monthly processing volume when it decided to build embedded payments. Before any vendor conversation, the team spent five weeks on program design: economics modeling at 12 and 36 months, program model evaluation, compliance framework design, bank partner criteria, and infrastructure requirements.
The economics model at 36-month volume ($8M monthly projected) showed: BaaS would generate approximately $540,000 annually at 75 bps net. A direct bank program would generate approximately $1.28M at 160 bps net. The gap — $740,000 annually — justified the bank relationship investment. The decision was to build toward a direct bank program from the start, accepting the 10-month build timeline rather than launching on BaaS and migrating later.
The bank diligence process took 8 months. The compliance infrastructure build took 10 weeks concurrent with diligence. The program launched at month 10 at $2.8M monthly — slightly below the $3M projection. Within 4 months of launch, the program was at $5M monthly and generating $64,000 in monthly payment economics. The build investment: $220,000 over 10 months.
Where they are today
Both companies at $8M monthly processing volume. Company A: direct program now live, generating approximately $128,000 monthly in payment economics ($1.54M annually at 160 bps net). Company B: direct program, same economics, same volume. The difference: Company B has been generating these economics since the program launched three years ago. Company A has been generating them for 14 months.
The cumulative economics difference over the three years since Company B launched: approximately $2.8M. Some of that difference was unavoidable — Company A needed time to reach the volume that justified the direct program. But the migration delay, the migration cost, and the 14 months of BaaS economics gap during migration account for approximately $1.1M of recoverable difference at current volume.
What made the difference
Company B did not have better payments expertise than Company A. It had a conversation about program design before the vendor conversation — one that surfaced the economics math, the migration cost, and the build timeline in the sequence that allowed a deliberate decision rather than a reactive one.
Company A made a perfectly reasonable decision with the information it had when it launched on Stripe. The problem was not the Stripe decision — it was the absence of the 36-month economics model that would have made the tradeoff visible before the contract was signed.
The five weeks of architecture work that Company B did before vendor selection cost approximately $30,000–$40,000 in internal time. The return on that investment, measured against Company A's counterfactual, was approximately $1.1M over three years — a 27x return on the architecture investment.
Which company does your payments program look like? The leakage calculator can estimate your current economics gap, or talk with us about where your program is and what the path forward looks like.